Despite its modest and variable incidence in the general population (ranging from 3 to 20 per 1,00010,11), ONHD remains a clinically important entity for several reasons. Primarily, ONHD have been a well-documented source of visual field defects, occurring in approximately 24% to 87% of involved eyes.12,13 Moreover, recent retrospective studies also suggest that these defects are progressive as the morphology of disc drusen evolves with age.10,14 Secondarily, ONHD simulate the nerve fiber layer loss and visual field defects seen in glaucomatous (and other) optic neuropathies. Thus, it proves to be a considerable confounder in the management of other progressive or comorbid eye conditions. Third, the mechanical role of ONHD as a risk factor for ischemic optic neuropathy15,16 and macular causes of vision loss (subretinal neovascularization, hemorrhages, and serous retinal detachments)17–19 has been observed (although no reports document it as a statistically significant covariant). ONHD are also found in the rare disorder pseudoxanthoma elasticum in association with angioid streaks.20
The first objective measurement we propose for monitoring ONHD progression is the disc size (as determined by RPE termination). In our cohort, 6 of 9 eyes demonstrated a cross-sectional disc diameter (as measured by this technique) greater than the diameter appreciated on the en face SLO image. Because measurement features are not available for the SLO image, statistical significance for this difference could not be evaluated. Nonetheless, this finding is consistent with a recent study reporting a significantly larger disc size on OCT cross-sectional imaging than on digital and OCT fundus image, attributed to the overestimation caused by artifactual shadowing of the RPE by overlying vasculature or peripapillary atrophy.21 As such, cross-sectional images were selected to traverse the geometric center of the disc and ONHD, but to avoid the passage of rim vessels. Although marginal drusen may also shadow adjacent RPE and overestimate disc size, a progressive increase in this measurement on a similarly oriented B-scan may indicate morphologic progression of the drusen.
Note that for all control group eyes, the disc size as determined by RPE termination on the OCT B-scan was not larger than the corresponding disc SLO image. This difference can be attributed to the fact that in the absence of marginal lesions or blood vessels, the scleral canal is accurately registered without obscuration caused by optical shadowing. Also, the location of a druse within the nerve head and relative proximity to the disc margin influences the measured disc size based on RPE termination. Similarly, the presence of peripapillary atrophy in normal eyes can also affect measurement of disc size. Therefore, a consistent difference in the RPE-based disc size cannot be expected between eyes with ONHD and normal eyes.
In similar fashion, we propose the serial measurement of posterior shadowing to assess lateral expansion of the drusen. A third parameter to reflect a progressive fullness or obliteration of the cup is the relative position of the base of the cup to an arbitrary reference point. In our study, the most consistent, easily discernible, and reproducible reference point was the level of the pigment epithelium. Although this level was generally disparate on either side of the nerve head, for reference, the side with the flattest registration of the pigment epithelium was assigned. As expected, the central cup was more shallow and ill defined in our study patients compared to 12 optic discs in normal healthy volunteers.
Retinal thickness maps and coronal scans (C-scans) have obvious utility in the study of retinal and macular pathology22; however, for the imaging of the nerve head, ONHD were poorly delineated on C-scans, likely due to the rapid change in contour along the z-axis. Thus, we do not propose C-scan parameters of the nerve head to follow morphologic changes associated with ONHD.
The frequency of visual field defects in adults with ONHD has been reported to range from 24% to 87%, with the highest rate of occurrence in eyes with superficial drusen.12,13 Visual field loss was noted in only 1 of 9 (11%) eyes in this study. This disparity can be likely attributed to the small sample size. With a larger cohort and a long-term investigation, a higher incidence and progression of visual field loss would be expected.
Additional studies have recently been published demonstrating OCT-based characteristics of ONHD. Wester et al. compared morphologic characteristics of optic nerves from patients with ONHD, papilledema secondary to idiopathic intracranial hypertension, and small optic nerves. These observations were made with the Stratus and Cirrus spectral-domain OCT (Carl Zeiss Meditec, Dublin, CA). In addition to elevation of the optic nerve surface, ONHD appear as circular optically empty spaces bounded by an anterior and a less apparent posterior line of reflectance. Eyes with papilledema demonstrated a more consistent elevation of the nerve with a highly reflective anterior boundary; optically empty cavities were not noted. Small nerves demonstrated central elevation without any discrete boundaries of reflectance, whether anterior or posterior.23
Yi et al. investigated the use of a prototype spectral-domain OCT to calculate the location, shape, and volumes of ONHD with a new algorithm. In turn, the volumes were correlated with visual field mean deviation. This study also confirmed that ONHD are signal-poor regions within the optic nerve head with relatively high-signaled borders. ONHD volume was also positively correlated with the visual field mean deviation values.24
A study conducted by Johnson et al. compared 20 eyes with ONHD versus 20 eyes with optic disc edema and 20 normal eyes.25 Qualitatively, they observed that eyes with ONHD had “lumpy-bumpy” internal contours, whereas eyes with optic nerve edema had smooth internal contours. Quantitatively, a peripapillary subretinal hyporeflective space was identified in both groups. However, in eyes with optic nerve edema, this hyporeflective space was observed to be thicker at a greater radius (2.0 mm) from the center of the disc, whereas in eyes with ONHD, this space rapidly tapered off. Additionally, the nasal retinal NFL thickness was found to be greater than in eyes with ONHD.
There are several limitations of the current study. The first concern is the operator-dependent selection of the cross-sectional B-scan. Given the irregular three-dimensionality of ONHD, it is conceivable that differently oriented sections may overestimate or underestimate lesion size. Also, the authors recognize that a marginal drusen shadow may obscure the RPE tip. However, because the main point of interest is relative change over time rather than the absolute value at a single point in time, we believe that the selection of the B-scan can be nonspecific, provided that the geometric center is registered while avoiding marginal blood vessels. Accordingly, the second concern is the ability to reproduce the scan for longitudinal monitoring. This depends largely on the operator’s ability to select a similarly oriented B-scan. With close attention to disc anatomy (as demonstrated on the corresponding SLO image), it has been our experience that the same or a similarly oriented scan can be identified for purposes of measurement. Of note, with the latest version of the OCT/SLO system, the spectral-domain algorithm offers multi-examination analysis of progression and/or regression at each point of the topographic comparison maps. With the user selecting the baseline examination for the comparison, the analysis generates reports with thickness and volume changes over time. This automatic selection of corresponding scans obviates the concerns of manual selection and erroneous tracking in our proposed parameters.
We also acknowledge the relatively small sample size in our investigation. Because our primary intention of this pilot study was to survey the application of this technology to a known pathology, we hope that a larger sample of patients would help determine statistically significant differences compared to normal optic discs or changes over time (using the newer spectral-domain models).
Our series illustrates the high-resolution capabilities of the OCT/SLO technology in contrast to standard techniques. Although ultrasonography offers the highest sensitivity for the detection of ONHD, the possibility of missing small lesions exists as expertise and experience of the ultrasonographer become important.1 The combined OCT/SLO offers improved visualization of ONHD size and morphological changes of adjacent retinal structures. With the advances of spectral-domain registration and improved light-source technology, the multiple transverse scans of high axial resolution will likely result in increased sensitivity in detection of both buried and visible lesions. Although a larger, prospective study may be required to compare the capabilities of OCT/SLO to other imaging techniques, this study found that OCT/SLO revealed unique and clinically helpful views of ONHD. With further, long-term studies documenting disc diameter, width of posterior shadowing secondary to drusen, and relative height of the optic cup, the OCT/SLO may help answer previously studied questions of ONHD pathophysiology. For example, temporal progression of buried to visible morphology and critical points in location and size resulting in NFL damage.